mnkbcs22021/whisper-small-mar

Whisper_Small_mar

This model is a fine-tuned version of openai/whisper-small on the openslr dataset. It achieves the following results on the evaluation set:

• Loss: 0.1866
• Wer: 26.4000

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 0.0001
• train_batch_size: 32
• eval_batch_size: 16
• seed: 42
• optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 50
• training_steps: 50
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
No log	0.2222	10	0.5078	64.3733
No log	0.4444	20	0.2853	37.3333
0.5228	0.6667	30	0.2131	30.3467
0.5228	0.8889	40	0.2020	32.64
0.183	1.1111	50	0.1866	26.4000

Framework versions

• Transformers 4.49.0
• Pytorch 2.6.0+cu124
• Datasets 3.3.2
• Tokenizers 0.21.0